If I was in India, and walked into a company meeting with a diverse set of people, I would assume the Indian looking person is the CEO. If I was in Japan, I would assume the Japanese looking person was the CEO.
This isn't racism. This is the human brain using pattern recognition to evaluate a situation and infer information as best it can under the circumstances.
So. One immediate problem, from statistics alone, is that the people you are meeting with are not drawn from a random sampling of the population. There is clearly a 50% chance that one of those two unknown people is the CEO. (If it were 3 people, a 33.33% chance, 4 people a 25% chance, and so on.) So I don't think your reasoning is sound.
Also, the end result of your argument is "the black person in this group isn't likely to be the CEO because they are black," and, well. I get that you're trying to arrive there through what you think is pure mathematics, but it is nonetheless the kind of conclusion that I believe the kids call "problematic."